King’s Research Portal

DOI:10.1371/journal.pone.0155475

Document VersionPublisher's PDF, also known as Version of record

Link to publication record in King's Research Portal

Citation for published version (APA):Sarkar, S., Dell'Acqua, F., Walsh, S. F., Blackwood, N., Scott, S., Craig, M. C., ... Murphy, D. G. M. (2016). Awhole-brain investigation of white matter microstructure in adolescents with conduct disorder. PL o S One ,11(6), [e0155475]. 10.1371/journal.pone.0155475

Citing this paperPlease note that where the full-text provided on King's Research Portal is the Author Accepted Manuscript or Post-Print version this maydiffer from the final Published version. If citing, it is advised that you check and use the publisher's definitive version for pagination,volume/issue, and date of publication details. And where the final published version is provided on the Research Portal, if citing you areagain advised to check the publisher's website for any subsequent corrections.

General rightsCopyright and moral rights for the publications made accessible in the Research Portal are retained by the authors and/or other copyrightowners and it is a condition of accessing publications that users recognize and abide by the legal requirements associated with these rights.

•Users may download and print one copy of any publication from the Research Portal for the purpose of private study or research.•You may not further distribute the material or use it for any profit-making activity or commercial gain•You may freely distribute the URL identifying the publication in the Research Portal

Take down policyIf you believe that this document breaches copyright please contact librarypure@kcl.ac.uk providing details, and we will remove access tothe work immediately and investigate your claim.

Download date: 18. Feb. 2017

RESEARCH ARTICLE

AWhole-Brain Investigation of White MatterMicrostructure in Adolescents with ConductDisorderSagari Sarkar1,2*, Flavio Dell’Acqua1,3, Seán Froudist Walsh5, Nigel Blackwood1,4,Stephen Scott6, Michael C. Craig1,4, Quinton Deeley1,4☯, Declan G. M. Murphy1,4☯

1 King’s College London, Sackler Institute for Translational Neurodevelopment and the Department ofForensic and Neurodevelopmental Sciences, Institute of Psychiatry, Psychology and Neuroscience, London,United Kingdom, 2 King’s College London, Department of Neuroimaging, Institute of Psychiatry, Psychologyand Neuroscience, London, United Kingdom, 3 King’s College London, Natbrainlab, Department ofNeuroimaging, Institute of Psychiatry, Psychology and Neuroscience, London, United Kingdom, 4 King’sCollege London, NIHR Biomedical Research Centre for Mental Health at South London and Maudsley NHSFoundation Trust and Institute of Psychiatry, Psychology and Neuroscience, London, United Kingdom,5 Friedman Brain Institute, Icahn School of Medicine at Mount Sinai, New York, New York, United States ofAmerica, 6 King’s College London, Department of Child and Adolescent Psychiatry, Institute of Psychiatry,Psychology and Neuroscience, London, United Kingdom

☯ These authors contributed equally to this work.* Sagari.Sarkar@kcl.ac.uk

Abstract

Background

The biological basis of severe antisocial behaviour in adolescents is poorly understood. We

recently reported that adolescents with conduct disorder (CD) have significantly increased

fractional anisotropy (FA) of the uncinate fasciculus (a white matter (WM) tract that connects

the amygdala to the frontal lobe) compared to their non-CD peers. However, the extent of

WM abnormality in other brain regions is currently unclear.

Methods

We used tract-based spatial statistics to investigate whole brain WMmicrostructural organi-

sation in 27 adolescent males with CD, and 21 non-CD controls. We also examined relation-

ships between FA and behavioural measures. Groups did not differ significantly in age,

ethnicity, or substance use history.

Results

The CD group, compared to controls, had clusters of significantly greater FA in 7 brain

regions corresponding to: 1) the bilateral inferior and superior cerebellar peduncles, cortico-

pontocerebellar tract, posterior limb of internal capsule, and corticospinal tract; 2) right

superior longitudinal fasciculus; and 3) left cerebellar WM. Severity of antisocial behavior

and callous-unemotional symptoms were significantly correlated with FA in several of these

regions across the total sample, but not in the CD or control groups alone.

PLOS ONE | DOI:10.1371/journal.pone.0155475 June 6, 2016 1 / 16

a11111

OPEN ACCESS

Citation: Sarkar S, Dell’Acqua F, Froudist Walsh S,Blackwood N, Scott S, Craig MC, et al. (2016) AWhole-Brain Investigation of White MatterMicrostructure in Adolescents with Conduct Disorder.PLoS ONE 11(6): e0155475. doi:10.1371/journal.pone.0155475

Editor: Pew-Thian Yap, University of North Carolina,UNITED STATES

Received: July 22, 2014

Accepted: April 29, 2016

Published: June 6, 2016

Copyright: © 2016 Sarkar et al. This is an openaccess article distributed under the terms of theCreative Commons Attribution License, which permitsunrestricted use, distribution, and reproduction in anymedium, provided the original author and source arecredited.

Data Availability Statement: All relevant data arewithin the paper and its Supporting Information files.

Funding: This work was funded by a privatedonation (to DGMM), together with infrastructuresupport from the Mortimer D. and Theresa SacklerFoundation (to DGMM), the Medical ResearchCouncil (MRC, United Kingdom), AIMS Network(G0400061/69344, DGMM, principle investigator), anongoing MRC funded study of brain myelination inneurodevelopmental disorders (G0800298/87573)(toDGMM), and the National Institute Health Research(NIHR) Biomedical Research Centre for Mental

Conclusions

Adolescents with CD have significantly greater FA than controls in WM regions correspond-

ing predominantly to the fronto-cerebellar circuit. There is preliminary evidence that varia-

tion in WMmicrostructure may be dimensionally related to behaviour problems in

youngsters. These findings are consistent with the hypothesis that antisocial behaviour in

some young people is associated with abnormalities in WM ‘connectivity’.

IntroductionConduct disorder (CD) is defined by a persistent display of antisocial behaviour such as decep-tion, theft, vandalism and violence within a 6–12 month period in under-18s [1], and occurs inup to 16% of school aged children [2]. Children with severe CD cost society 10 times more tosupport into adulthood than those without CD [3]. Further, CD is strongly associated withother mental health problems (e.g. substance abuse [4] and mood disorders [5]), and antisocialpersonality disorder (ASPD) as adults [6]. It is likely that CD arises from a complex constella-tion of factors–and cannot be simply explained by any one putative social or biological causa-tive agent alone. Nevertheless, despite the significant impact of CD on affected individuals andsociety as a whole, its biological determinants are still poorly understood.

Attempts to identify the neurobiological bases of CD using in vivo brain imaging havereported abnormalities in the structure [7–12] and function [13–18] of temporo-limbic andprefrontal brain regions. Hence there is increasing evidence that specific brain regions may beimplicated in CD. However, brain regions do not act in isolation–they form part of large scaleneural networks. Thus it is important to also examine the ‘connectivity’ of particular neuralsystems.

There is preliminary evidence that antisocial behaviour is associated with functional differ-ences in the limbic-prefrontal network (that is associated with the generation of complex socialand emotional behaviours) [15, 18–20]. The anatomical substrate for these functional differ-ences in neural networks is unknown. However, we can now address this issue as the micro-structural organisation of white matter (WM) tracts connecting neural systems can be indexedby measuring their fractional anisotropy (FA) using diffusion tensor magnetic resonance imag-ing (DT-MRI). FA is an index that quantifies directional differences in the diffusion of watermolecules inside tissues. FA values range from 0 (perfectly isotropic diffusion) to 1 (perfectlyanisotropic diffusion)—providing a proxy measure of tissue microstructural organization [21].The microstructural basis for FA value is thought to lie with properties such as the organisationwithin and between fibres, axonal diameter, and myelination [22, 23].

We recently reported that adolescents with CD have increased FA of the uncinate fasciculus(UF) a major limbic-prefrontal WM connection, as compared to non-CD controls [24]. How-ever, we used tractography based methods and so only examined a predefined tract of interest,and not whole brain. Therefore, the regional specificity of our previous finding (i.e. whetherany additional tracts show abnormal microstructural diffusion properties in CD) is unknown.Tract-based spatial statistics (TBSS) is an automated method of whole-brain voxel based WManalysis [25].

Several studies of antisocial adults and children have examined whole brain WM usingTBSS. For example, Sundram et al [26] reported reduced FA in the corpus callosum, coronaradiata, inferior fronto-occipital fasciculus (IFOF), uncinate fasciculus, and internal capsule inmales with ASPD. Similarly, Hoppenbrouwer et al’s [27] study of psychopathic offendersreported reduced FA in areas including the uncinate fasciculus, IFOF, and anterior thalamic

Whole Brain Conduct Disorder

PLOS ONE | DOI:10.1371/journal.pone.0155475 June 6, 2016 2 / 16

Health at King’s College London, Institute ofPsychiatry and South London & Maudsley NHSFoundation Trust (to DGMM). The funders had norole in study design, data collection and analysis,decision to publish, or preparation of the manuscript.

Competing Interests: The authors have declaredthat no competing interests exist.

radiation. However, studies of developmental samples have produced discrepant findings,reporting FA to be increased [28, 29], decreased [30] or no different [31] in adolescents withCD compared to controls.

The discrepancies between these studies may lie with methodological differences, includingsmall sample size [31], and mixed sex samples [30, 31]. Specifically, in typical children and ado-lescents WM develops at a greater rate in males than females [32], so the increased FA we andothers have observed in boys with CD may indicate an exaggeration of typical patterns of WMdevelopment [24, 28, 29, 33]. Furthermore, recent studies highlight important sex differencesin the strength and distribution of WM ‘connectivity’ between sexes [34]. Our study sought toclarify the nature and extent of whole brain WM differences in CD using TBSS using a large,well characterised sample of adolescent males with CD and a group of non-CD controls.

Methods and MaterialsThis study was approved by the Joint South London and Maudsley Research Ethics Committee(243/00).

ParticipantsTwenty seven participants with CD aged between 12 and 19 years were recruited as part of alarger study (see [24]) from: (i) a Kings College London, Institute of Psychiatry database ofadolescents with conduct problems; (ii) three Youth Offending Teams; (iii) five Pupil ReferralUnits (facilities providing education to children who cannot attend mainstream schools, e.g.following school exclusion); (iv) four youth projects; and (v) two mainstream educational insti-tutions. A further twenty-one right handed males were recruited as controls from the generalpublic, through schools and youth services (i.e. youth clubs, ‘Connexions’, and several youthcharities) within the same geographical areas (deprived and inner city) as the CD group.Groups did not significantly differ in age, ethnicity, and self-reported history of alcohol or can-nabis use. Furthermore, measures of current hyperactivity did not differ significantly betweengroups (see Table 1) and each group contained an equal number of boys with a prior diagnosisof ADHD (n = 2). To check for comorbid conditions, participants and parents were inter-viewed to ensure participants had no previous psychiatric diagnoses.

All study participants: satisfied MRI safety requirements and were medication free, did nothave psychiatric or substance use disorders (other than CD, ADHD, or referrals for anger man-agement), spoke English as their first language, and were right handed as assessed by the Edin-burgh Handedness Inventory [35]. IQ was measured using the vocabulary and matrixreasoning subtests of the Wechsler Abbreviated Scale of Intelligence (WASI; [36]). Weexcluded individuals with IQ<80. Controls had a significantly higher IQ than CD individuals.Hence we co-varied for IQ in all subsequent analyses (see S1 Dataset).

MeasuresQuestionnaires. Parent and self-report versions of the Strengths and Difficulties Ques-

tionnaire (SDQ; [37]) and Antisocial Process Screening Device (APSD; [38]) were adminis-tered. The SDQ was used to obtain conduct problem, emotional problem, and hyperactivitymeasures, while the APSD assessed callous-unemotional (CU) traits, narcissism, and impulsiv-ity. Following methods of other groups, accepted subscales for both measures comprised thehigher rater’s score for each item [14].

Interviews. CD and Oppositional defiant disorder (ODD) subsections of the KiddieSchedule for Affective Disorders and Schizophrenia—Present and Lifetime version(K-SADS-PL; [39]) were used to obtain a research diagnosis of CD. Screening interviews for

Whole Brain Conduct Disorder

PLOS ONE | DOI:10.1371/journal.pone.0155475 June 6, 2016 3 / 16

these disorders were administered to all participants, with those meeting criteria for CD orODD given complete interviews for both disorders. No participants met criteria for ODD inthe absence of CD. Finally, in order to assess psychopathic traits participants meeting CD crite-ria who additionally scored� 20 on the APSD parent or self-report questionnaire were theninterviewed using the Psychopathy Checklist Youth Version (PCL-YV; [40]). Scores of� 20were used to indicate the presence of psychopathic traits [41]. Interviews were conducted by aresearch psychologist (SSarkar) trained and supervised by a psychiatrist (QD). Additionalinformation about antisocial behaviour was gathered from teachers, youth club workers, socialworkers and parents

Participants in the CD group had a history of serious aggressive and violent behaviour,including: robbery, burglary, grievous bodily harm, and sexual assault.

ProcedureFull written informed consent was taken from participants, and additionally from a parent/guardian where boys were under 16 years old.

DT-MRI acquisition. Each DT-MRI image was acquired using a GE Signa HDx 3.0T MRscanner (General Electric, Waukshua, WI, USA), with actively shielded magnetic field gradi-ents (maximum amplitude 40 mT m-1). The body coil was used for RF transmission, and an 8channel head coil for signal reception, allowing a parallel imaging (ASSET) speed up factor oftwo. Head movement was minimised by fitting extra padding beside participants’ heads. Eachvolume was acquired using a multi-slice peripherally-gated doubly refocused spin echo EPIsequence, optimised for precise measurement of the diffusion tensor in parenchyma, from 60contiguous near-axial slice locations with a voxel size of 1.875 x 1.875 x 2.4 mm. The echo timewas 104.5 ms while the effective repetition time varied between subjects in the range 12 and 20

Table 1. Group characteristics. FSIQ—Full Scale Intelligence Quotient; SDQ–Strengths and Difficulties Questionnaire; APSD–Antisocial Process Screen-ing Device; SD–standard deviation; #Excluding alcohol.

Conduct disorder (n = 27) Mean (SD) Healthy controls (n = 21) Mean (SD) P value

Age in years 16 (2) 16 (2) 0.99

Mean FSIQ 99 (8) 110 (15) 0.01*

Conduct problems (SDQ) 6 (2) 3 (1) 0.00**

Hyperactivity (SDQ) 7 (2) 6 (2) 0.20

Emotional problems (SDQ) 2 (2) 3 (2) 0.15

Total problems (SDQ) 19 (5) 13 (4) 0.00**

Callous-unemotional traits (APSD) 7 (2) 5 (2) 0.00**

Total score (APSD) 25 (7) 18 (6) 0.00**

Ethnicity (%) Chi2

White 52 62 0.49

Black/African-Caribbean 33 24 0.47

Other 15 14 0.96

Substance use (%) n = 20 n = 20 Chi2

Cannabis—ever used 60 45 0.34

Cannabis–past month 30 35 0.74

Alcohol–ever used 75 95 0.08

Alcohol–past month 55 65 0.52#Any other drug–ever used 44 43 0.09

*significant p value <0.05

**significant p value p<0.01

doi:10.1371/journal.pone.0155475.t001

Whole Brain Conduct Disorder

PLOS ONE | DOI:10.1371/journal.pone.0155475 June 6, 2016 4 / 16

RR intervals. Based on the recommendations of Jones et al [42], the maximum diffusionweighting was 1300 s mm-2, and at each slice location, 4 images were acquired with no diffu-sion gradients applied, together with 32 diffusion-weighted images in which gradient directionswere uniformly distributed in space. The sequence ran for approximately 15 minutes.

DT-MRI data pre-processing. All data were first converted to NIFTI format and theneach raw diffusion dataset underwent a full quality control check where all B0s and DiffusionWeighted volumes were visually inspected for image corruption, motion artefacts, and signaldrop-out effects using the light-box function available inside fslview (fmrib Software Library,www.fmrib.ox.ac.uk/fsl). Datasets showing more than 2 motion artefacts in different volumeson the same slice were removed from the study. Datasets showing significant head movements(>1 cm) were removed. No participant data acquired in this study required removal due tomotion artefacts. Data were eddy current and motion corrected using Explore-DTI [43]. Thediffusion tensor was estimated following removal of outlier data (RESTORE function [44]),and FA and MDmaps were generated. Full details are given elsewhere [42].

DT-MRI analysis and statistics. First, the FAmaps were transformed into standard stereo-tactic space using a study specific template generated in FSL (www.fmrib.ox.ac.uk/fsl), which wasthe FAmap most representative of all FA images within the sample. All FA maps were averagedinto a mean FAmap for the whole sample and an average skeleton was created onto which eachparticipant’s aligned FA data was projected. Finally, TBSS (part of FSL [45]) was applied to diffu-sion data for voxelwise analysis of whole brain WM [25]. Age and FSIQ, by group, were includedas covariates in the design matrices used in the analysis, and results were corrected for multiplecomparisons by FSL. The relevant contrasts identified regions where: FA in CD>Controls; FA inControls>CD; FA in CD positively correlated with age; and FA in controls positively correlatedwith age. Regions showing significant FA differences (with a threshold of p<0.05; corrected formultiple comparisons) between groups were identified with reference toWM atlases [46]. Forlater correlation analysis (with behavioural and age data) it was necessary to extract FA valuesfrom regions of interest using masks created in FSL using the John Hopkins University (JHU)WM atlas [47]. FA values were correlated with SDQ and APSD behavioural scores, using Spear-man’s rho non-parametric correlations; correlations were Bonferroni corrected (see S1 Dataset).

Results

Between group analysisThe CD group had significantly greater FA as compared to non-CD controls in 7 regions: 1)bilaterally in the inferior and superior cerebellar peduncles, corticopontocerebellar tract, poste-rior limb of internal capsule, and corticospinal tract; 2) in the right superior longitudinal fascic-ulus; and 3) in left cerebellar WM (see Fig 1). Controls had no areas with significantly greaterFA in comparison to the CD group.

Behavioural relationshipsThere was no significant correlation between age and FA in either group, and there were no sig-nificant correlations between behaviour and FA within the CD or non-CD control groups alone.

In the whole sample, however, SDQ and APSD scores were positively correlated with FAvalues in all of the 7 regions that showed between group differences (see Table 2 and Fig 2).

DiscussionWe used whole-brain voxel-based DT-MRI to explore WMmicrostructural diffusion proper-ties in a sample of adolescent boys with CD and a non-CD comparison group. There was

Whole Brain Conduct Disorder

PLOS ONE | DOI:10.1371/journal.pone.0155475 June 6, 2016 5 / 16

significantly increased FA within regions corresponding with the trajectories of several WMtracts in CD boys compared to non-CD controls, namely: 1) bilaterally in the inferior and supe-rior cerebellar peduncle, corticospinal tract, internal capsule (posterior limb), and corticopon-tocerebellar tract; 2) in the right superior longitudinal fasciculus; and 3) in the left cerebellarWM. There were no areas of significantly reduced FA in CD as compared to controls. Therewere significant positive correlations between FA and behavioural variables in all 7 tractregions showing between group differences in the total sample, but not in the CD or controlgroups alone.

Firstly, it is important to note that our analysis did not identify between group differences inlimbic tracts such as the uncinate fasciculus (UF), as was the case in our group’s previousDT-MRI tractography study [24]. The reason for this may arise from the nature of TBSS analy-sis. The software used to perform the voxel-wise statistics (Randomise) sequentially comparesevery voxel to its corresponding voxel within each participant’s brain. Thus, differences seenusing TBSS can be viewed as more locally defined (i.e. clusters of voxels) than tractography. Incontrast, tractography can be used to detect more subtle differences along specific tracts. Thus,while we reported increased FA in the UF compared to the non-limbic control tracts in ourprior study, this is likely not as pronounced in individual groups of voxels as the increases inFA found in the projection tracts observed in this study.

The increased FA we report contrasts with findings from studies of antisocial adults (e.g.[48]), possibly resulting from abnormal/precocious WM development in CD during childhood,which plateaus later in development [29]. This hypothesis is consistent with similar patterns ofWMmaturation observed in other specific neurodevelopmental disorders. For instance, struc-tural MRI studies show that compared to neurotypical controls, people with autistic spectrumconditions (ASC) show increased WM volume in early childhood, but the opposite pattern (i.e.decreased volume) during adolescence [49]. This early acceleration also corresponds with

Fig 1. Regions of significantly increased fractional anisotropy in conduct disordered adolescentscompared to healthy controls. Key: R-right; L- left; A-anterior; P-posterior; green indicates mean FA (fractionalanisotropy) skeleton; red denotes areas of significantly greater (p < .05) FA in CD in: (i) bilateral superior cerebellarpeduncle; (ii) left cerebellar white matter; (iii) right superior longitudinal fasciculus; (iv) bilateralcorticopontocerebellar tract; (v) bilateral posterior limb of internal capsule; (vi) bilateral inferior cerebellar peduncle;(vii) bilateral corticospinal tract; (viii) bilateral corticopontocerebellar tract

doi:10.1371/journal.pone.0155475.g001

Whole Brain Conduct Disorder

PLOS ONE | DOI:10.1371/journal.pone.0155475 June 6, 2016 6 / 16

DT-MRI research showing FA to be increased in specific brain regions (e.g. right inferior fron-tal gyrus and left occipital lobe [50]; corpus callosum and cingulum [51] in ASC. Future longi-tudinal studies should be undertaken to clarify whether this mechanism also occurs in CD.

However, that our study revealed FA to be increased, rather than decreased in the CD groupis consistent with Zhang et al’s [28] study, and our group and others’ previous tractographystudy findings [24, 29]. Zhang et al reported increased FA in the corpus collosum and bilateralcorona radiata. Although we did not find FA differences in the CC, the corticopontocerebellarand corticosprinal tracts in which we found FA increases, project through the corona radiataon their path from the cortex [52]. Therefore, while the precise location of FA increases differsbetween the two studies, both results may point to FA increases in the same tract. Finally, ourresults are in line with findings of increased FA in frontal WM associated with greater risk tak-ing in young people [53]. The authors proposed that early WMmaturation may stem fromgreater number/type of life experiences.

Although the precise significance of FA is not agreed on, it is regarded as a measure of inter-and intra-axonal properties, including the organisation within and between fibres, axonaldiameter, and myelination [22, 23]. Therefore, the increased FA reported here most likelyreflects differences in WM organisation or greater myelination of axonal tracts in CD. Alterna-tively, it is possible that rather than acceleration of the progressive process of myelination,increased FA in CD may reflect dysfunction in axonal pruning, the process by which surplusneuronal processes laid down in earlier developmental stages are removed, thereby refining

Table 2. Correlations between SDQ and APSD scores and fractional anisotropy in whole sample. JHU–John Hopkins University; SDQ–Strengths andDifficulties Questionnaire; APSD–Antisocial Process Screening Device; CU–callous-unemotional; r–Spearman’s correlation coefficient; p–two-tailed signifi-cance level.

SDQ APSD

JHU white matter atlas region Total problems Conductproblems

Totalproblems

CU traits Impulsivity

r p r p r p r p r p

Corticospinal tract

Left 0.48 .00* 0.41 .00* 0.21 .15 0.19 .20 0.12 .40

Right 0.48 .00* 0.41 .00* 0.28 .06 0.23 .11 0.21 .14

Internal capsule

Left posterior limb 0.33 .03 0.42 .00* 0.28 .06 0.34 .02 0.26 .08

Right posterior limb 0.43 .00* 0.41 .00* 0.30 .04 0.44 .00* 0.28 .05

Cerebellar peduncle

Left superior 0.48 .00* 0.42 .00* 0.22 .14 0.20 .17 0.13 .40

Right superior 0.44 .00* 0.36 .01 0.26 .07 0.23 .12 0.18 .22

Left inferior 0.29 .05 0.36 .01 0.31 .03 0.24 .11 0.25 .09

Right inferior 0.38 .01 0.38 .01 0.31 .03 0.22 .14 0.37 .01

Corticopontocerebellar tract

Left 0.43 .01 0.45 .00* 0.27 .06 0.29 .05 0.23 .12

Right 0.50 .00* 0.40 .01 0.30 .04 0.30 .04 0.27 .07

Superior longitudinal fasciculus

Right 0.31 .03 0.25 .09 0.33 .02 0.32 .03 0.28 .06

Cerebellar white matter

Left 0.35 .02 0.50 .00* 0.32 .03 0.36 .01 0.21 .15

*significant at p<0.05 after Bonferroni correction

doi:10.1371/journal.pone.0155475.t002

Whole Brain Conduct Disorder

PLOS ONE | DOI:10.1371/journal.pone.0155475 June 6, 2016 7 / 16

information processing. However, it was not possible to directly test these two competingexplanations in this study–but this issue could be addressed using in vivomyelin mapping [54].

The clusters of significantly increased FA we observed in the CD group correspond predom-inantly to projection tracts, which connect cortical and subcortical regions, including: the corti-cospinal and corticopontocerebellar tracts, and the posterior limb of the internal capsule(which carries the corticospinal tract fibres [55]). These WM tracts connect the cerebral cortexwith the brainstem and the pons (which is further connected to the cerebellum). As well as itsrole in sleep and respiration, functions of the pons include sensory analysis, motor control, andthe production of facial expressions. Animal studies demonstrate that pontine stimulation canprovoke predatory attack [56]. This may explain how significant WM differences (which couldcause differences in signal conduction through this region) may be associated with aggressivebehaviour in CD.

The tracts within the prefrontal-thalamic-cerebellar circuit that we found to be abnormal inCD are also abnormal in both adolescents [57] and adults with alcohol use disorders [58–60].These disorders, alongside early alcohol use, are predicted by childhood antisocial behaviour[61, 62]. Moreover, it has been suggested that these disorders arise from a common pathway[61]. Therefore, disruption to the prefrontal-thalamic-cerebellar WM pathways in CD may berelated to, or underlie, the structural abnormalities seen in AUDs. Alternatively, the abnormali-ties we found in CD may be confounded by differences in alcohol use. However, this is unlikelyto fully explain the differences we found, as the CD and control samples in our study did not

Fig 2. Correlations between FA and behavioural measures in whole sample (above) and by group (below).There were significant correlations between: 1) SDQ total problems and FA of the bilateral corticospinal tract, posteriorlimb of the internal capsule, superior cerebellar peduncle, and corticopontocerebellar tract; left cerebellar WM; and theright inferior cerebellar peduncle, and superior longitudinal fasciculus; 2) SDQ conduct problems and FA of the bilateralcorticospinal tract, posterior limb of the internal capsule, superior cerebellar peduncle, inferior cerebellar peduncle, andcorticopontocerebellar tract; and left cerebellar WM; 3) APSD total problems and FA of the bilateral inferior cerebellarpeduncle; left cerebellarWM; and the right posterior limb of the internal capsule, superior longitudinal fasciculus, andcorticopontocerebellar tract; 4) APSD callous-unemotional traits and FA of the right posterior limb of the internalcapsule, superior longitudinal fasciculus, corticopontocerebellar tract; and left cerebellar WM; and 5) APSD impulsivityand the right inferior cerebellar peduncle.

doi:10.1371/journal.pone.0155475.g002

Whole Brain Conduct Disorder

PLOS ONE | DOI:10.1371/journal.pone.0155475 June 6, 2016 8 / 16

significantly differ in self-reported alcohol and substance use. Thus, abnormalities in thesetracts in a) young people with alcohol use disorders and b) young people with antisocial behav-iour may reflect shared biological determinants affecting WMmicrostructure which (respec-tively) moderate risk of alcohol misuse and antisocial behaviour.

Cerebellar peduncleWe also found between group differences in FA in the superior and inferior cerebellar peduncles.These fibre bundles are composed of efferent projections to the thalamus (superior cerebellarpeduncle) or afferent projections from the spinal cord (inferior cerebellar peduncle). The majorinput to the cerebellar peduncle originates from the prefrontal, as opposed to the motor, cortex(PFC) [63], and this input is received in the most part via the corticopontine tract [64]. Thus, FAdifferences in these areas may indicate abnormal microstructure within theWM circuit connectingthe PFC to the pons and cerebellum in boys with CD. The cause of this is unknown–but may stemfrom abnormal input from the PFC–as a number of prior neuroimaging studies have reported thatCD individuals have significant differences in the anatomy and function of this brain region (e.g.[11, 41]). However, the current data cannot determine whether abnormality of the PFC is primaryor secondary to differences in this tract. Longitudinal studies are required to elucidate this issue.

Evidence that increased FA of the cerebellar peduncle may be relevant to the generation ofantisocial behaviour comes from patients with damage to PFC-cerebellar circuits. These individ-uals display some of the same deficits observed in antisocial children and adults. (i.e. emotionalprocessing difficulties [65–67] and deficits of conditional associative learning [68]). Further,abnormal microstructural organization of cerebellar tracts has been reported in other neurodeve-lopmental disorders with differences in social function; e.g. in people with Autism Spectrum Dis-orders and schizophrenia [69, 70]. Thus, we do not suggest that abnormalities in PFC-cerebellarconnections are specific to CD. Rather, they may underlie some aspects of social cognition defi-cits in a number of neurodevelopmental disorders. Nevertheless, these studies of other disordersonly reported abnormality in selective tracts within the cortical-cerebellar-thalamic-cortical net-work, or of increased FA in some tracts but decreases in others. For example, reduced FA of thesuperior cerebellar peduncle was reported in Asperger syndrome, in the absence of deficits in cer-ebellar input pathways [69]. Further, in people with schizophrenia, increased FA of the superiorcerebellar peduncle was found alongside reduced FA of the middle cerebellar peduncle, whichcontains afferent fibres [70]. In contrast, we found that individuals with CD have uniformlyincreased FA in all major cortical-cerebellar connections as compared with controls (i.e. the cere-bellar peduncle, corticopontocerebellar tract, corticospinal tract, cerebellar WM). This suggeststhat a more generalized cortico-cerebellar ‘dysconnectivity’ (i.e. ‘abnormal functional integrationof brain processes’ [71]) may underlie the emotional and behavioural deficits that characteriseCD. Several studies have reported that young people with CD have concurrent abnormality inboth prefrontal and cerebellar anatomy [8, 11, 12] or function [16, 18]. This suggests that boththese regions may be abnormal in CD and that it is abnormal interconnection between these tworegions that is of importance. Taken together, this study and that of others support the conjecturethat boys with CD have abnormal cortico-cerebellar ‘connectivity’. It is unknown, however,which comes first (i.e. differences in brain anatomy or function, or behaviour).

Corticospinal tractIndividuals with CD also had increased FA in the corticospinal tract. This tract contains pre-dominantly motor axons, yet some prior work also suggests that it may be involved in emotionprocessing. For example, others reported [72], using transcranial magnetic stimulation, that inhealthy controls threat signals, such as fearful facial expressions, increased the motor evoked

Whole Brain Conduct Disorder

PLOS ONE | DOI:10.1371/journal.pone.0155475 June 6, 2016 9 / 16

potential of the corticospinal tract significantly more than positive or neutral faces [72]. Thissuggests that the corticospinal tract interacts closely with the limbic system to coordinate themotor response to threat-related stimuli. In this context, the increased FA of this tract detectedby us in CD may indicate abnormality in responding to such cues. This suggestion fits withreports of abnormal neural responsivity to fearful faces and affective stimuli in both childrenwith CD/CU traits [14–17], and adults with ASPD and psychopathy [73–75]. Future studiescould investigate this relationship through correlating emotion processing measures, evokedpotentials, and FA of the corticospinal tract.

Superior longitudinal fasciculusThe area showing increased FA in CD corresponded with one of the subsections of the superiorlongitudinal fasciculus (SLF1), which connects the dorsal and medial parietal lobe with the dor-sal and medial frontal lobe [76]. The precise function of this tract is not clear, as in vivo investiga-tion of its anatomical characteristics has not long been possible [77]. However, the SLF1 isassociated with the updating of verbal, as well as spatial, information; and deficits in verbal intelli-gence are noted in children with CD [78]. Thus, increased FA in this tract in CDmay interferewith normal information processing and manifest as a verbal cognitive deficit. Investigationsusing DT-MRI tractography to correlate SFL1 integrity with verbal IQ scores may verify this.

Behavioural relationshipsIn addition to the between group differences in FA, our behavioural analysis suggests that theregions in which greater FA was reported in CD may contribute towards the generation of theemotional and behavioural features of this disorder. While antisocial behaviour measures andFA were not correlated within the CD group, there was a significant positive correlation in thesample as a whole. It is possible that this correlation is caused by the fact that we merged datafrom two different sources–and so is ‘driven’ simply by between group differences and doesnot reflect a true dimensional (as opposed to categorical) relationship. In contrast, the regionsof increased FA in CD may be dimensionally associated with antisocial behaviour across alladolescents, but the small sample size in this study may have prevented the correlations withineach group from reaching significance. Preliminary support for this suggestion is provided byFig 2, which exemplifies the distribution of SDQ and APSD scores against FA. This fits withthe idea that conduct problems may fall along a continuum, and that the relationship betweendifferences in brain and antisocial personality traits should be viewed as dimensional, ratherthan categorical [79].

Alternatively, however, it is possible that instead of FA differences underlying behaviouraldifferences, it is the behavioural differences themselves that drive differences in WMmicro-structure. Support for this suggestion comes from studies investigating how social and emo-tional experience can modulate FA. For example, children who experienced severe deprivationin childhood had significantly reduced FA of the left uncinate fasciculus as compared to con-trols [80]. Similarly, others reported that young adults exposed to high levels of parental verbalabuse in childhood have significantly decreased FA in two left hemisphere limbic tracts (cingu-lum and fornix) and the arcuate fasciculus [81]. Conversely, increased FA appears can resultfrom experience, such as after extensive practice of piano playing [82], meditation [83], andpossibly from activity-dependent myelination resulting from social experiences [84]. Togetherwith evidence of increased FA in high risk taking adolescents [53], it could be argued that theincreased FA observed in CD may reflect increased myelination occurring as a result of theseadolescents experiencing greater numbers/types of life experiences at an earlier age than theirtypical peers.

Whole Brain Conduct Disorder

PLOS ONE | DOI:10.1371/journal.pone.0155475 June 6, 2016 10 / 16

Therefore, the differences in tract integrity observed in developmental psychopathologies,such as in our study, most likely arise through a complex mixture of neurobiological, environ-mental and social factors. Future studies of WMmaturation should consider (for example) therelative contribution of social and environmental variables in order to evaluate the potentialrelevance of early interventions to moderate or prevent the course of developmental disorders.

LimitationsFirst, our groups showed a significant difference in mean FSIQ. Low IQ relative to the generalpopulation is characteristic of antisocial disorders [85, 86], so that cases and controls who areotherwise well matched may nevertheless significantly differ in IQ. We co-varied for FSIQ dur-ing our TBSS analyses to increase confidence that between group differences in FA are due todifferences in conduct problems rather than IQ. Equally, there does not appear to be a simplerelationship between IQ and FA in these tracts in adolescent boys (e.g. [87]).

Second, we did not find any correlations between FA and age in either group. This is in con-trast to evidence of increased FA in adolescents who score highly on a scale of ‘risk taking’ [53].A difference in sample size may explain these discrepant findings; therefore, future studies withlarger sample sizes may reveal a relationship between age andWM in CD.

Third, it should be noted that methodological factors may have contributed to the increasedFA values we detected. For example, DT-MRI derived outputs do not take into account complexWM organisation i.e. ‘crossing fibres’ [88, 89]. In the presence of multiple fibre orientations withinthe same voxel, the diffusion tensor provides only an average description of the real diffusion andthis could make the interpretation of FA changes more difficult. For example, within the right cor-ticospinal tract of Fig 1 (vii) there is an area that shows no significant FA difference flanked bytwo areas that do (corresponding to the corticospinal tract). It is possible either that the area show-ing no FA difference does not differ between groups, or, alternatively, that the two areas showincreased FA only because there are fewer fibres crossing these portions of the cortical spinal tract.

However, crossing fibres cannot explain the behavioural associations we observed in ourdata, as within both groups of subjects the same regions were associated with behaviouralscores. Nevertheless, possible methods for overcoming this limitation could be considered inthe future by using tract specific indices that are not/less affected by the presence of crossingfibres. (see [90, 91]), and using the latest methods of dealing with this issue [92]

Fourth, while our groups did not significantly differ in self-reported alcohol and cannabisuse it is widely noted that self-reported measures may be inaccurate. Furthermore, alcohol andsubstance use variables were coded for in a binary manner, which did not permit us to furthercontrol for these variables in our analyses. Future studies would benefit from assessing sub-stance use quantitatively and obtaining objective measures of alcohol/substance use (e.g. usinganalysis of hair samples). However, given that FA changes in white matter associated with alco-hol misuse tend to be diffuse reductions that affect multiple (mainly frontal) tracts suggeststhat substance misuse alone is unlikely to be driving our findings [93].

Last, it is important to note that while this study has identified correlations (i.e. associations)between brain structure and CD, it is not possible to clearly ascribe causation, define potentialbiological pathways, or rule out the influence of other factors on these neurodevelopmentaloutcomes. For example, data were not available that pertained to childhood trauma, negativefamily experiences, illnesses, and other events/neurochemical mechanisms that may influencebrain development throughout infancy and childhood.

In summary, this study examined whole brain WM in boys with CD. Adolescent males withCD have significantly greater FA than controls in WM regions corresponding to the fronto-cerebellar circuit. There is preliminary evidence that variation in WMmicrostructure may be

Whole Brain Conduct Disorder

PLOS ONE | DOI:10.1371/journal.pone.0155475 June 6, 2016 11 / 16

dimensionally related to behaviour problems in some youngsters. These findings are consistentwith the hypothesis that antisocial behaviour in young people is associated with abnormalitiesin WM ‘connectivity’ within the fronto-cerebellar network.

Supporting InformationS1 Dataset.(SAV)

AcknowledgmentsThis work was generously funded by a private donation; together with infrastructure supportfrom the Mortimer D and Theresa Sackler Foundation, the Medical Research Council (MRC,UK) AIMS Network (G0400061/69344, DGMM, principle investigator), an ongoing MRCfunded study of brain myelination in neurodevelopmental disorders (G0800298/87573), andthe NIHR Biomedical Research Centre for Mental Health at King’s College London, Instituteof Psychiatry and South London &Maudsley NHS Foundation Trust.

Author ContributionsConceived and designed the experiments: SS (first author) DM FD. Performed the experi-ments: SS (first author). Analyzed the data: SS (first author) FD SFW. Contributed reagents/materials/analysis tools: SS (first author) FD SFW. Wrote the paper: SS (first author) DM SS(fifth author) MCC NB QD.

References1. American PA. Diagnostic and Statistical Manual of Mental Disorders Fourth Edition—Text Revision.

2000.

2. Olsson M. DSM diagnosis of conduct disorder (CD)—A review. Nordic Journal of Psychiatry. 2008:1–11.

3. Scott S, Knapp M, Henderson J, Maughan B. Financial cost of social exclusion: Follow up study of anti-social children into adulthood. British Medical Journal. 2001; 323:191–4. PMID: 11473907

4. Kessler RC, Nelson CB, McGonagle KA, Edlund MJ, Frank RG, Leaf PJ. The epidemiology of co-occur-ring addictive and mental disorders: implications for prevention and service utilisation. Am J Orthopsy-chiatry. 1996; 66(1):17–31. PMID: 8720638

5. Vloet TD, Konrad K, Huebner T, Herpertz S, Herpertz-Dahlmann B. Structural and functional MRI- find-ings in children and adolescents with antisocial behavior. Behav Sci Law. 2008; 26(1):99–111. doi: 10.1002/bsl.794 PMID: 18327828

6. Gelhorn HL, Sakai JT, Price RK, Crowley TJ. DSM-IV conduct disorder criteria as predictors of antiso-cial personality disorder. Compr Psychiatry. 2007; 48(6):529–38. PMID: 17954138

7. Fairchild G, Passamonti L, Hurford G, Hagan CC, von dem Hagen EAH, van Goozen SHM, et al. Brainstructure abnormalities in early-onset and adolescent-onset conduct disorder. Am J Psychiatry. 2011;168(6):624–33. doi: 10.1176/appi.ajp.2010.10081184 PMID: 21454920

8. Huebner T, Vloet TD, Marx I, Konrad K, Fink GR, Herpertz SC, et al. Morphometric brain abnormalitiesin boys with conduct disorder. J Am Acad Child Adolesc Psychiatry. 2008; 47(5):540–7. doi: 10.1097/CHI.0b013e3181676545 PMID: 18356764

9. Sterzer P, Stadler C, Poustka F, Kleinschmidt A. A structural neural deficit in adolescents with conductdisorder and its association with lack of empathy. Neuroimage. 2007; 37(1):335–42. PMID: 17553706

10. Kruesi MJ, Casanova MF, Mannheim G, Johnson-Bilder A. Reduced temporal lobe volume in earlyonset conduct disorder. Psychiatry Res. 2004; 132(1):1–11. PMID: 15546698

11. De Brito SA, Mechelli A, Wilke M, Laurens KR, Jones AP, Barker GJ, et al. Size matters: increasedgrey matter in boys with conduct problems and callous-unemotional traits. Brain. 2009; 132(Pt 4):843–52. doi: 10.1093/brain/awp011 PMID: 19293245

12. Fahim C, He Y, Yoon U, Chen J, Evans AC, Perusse D. Neuroanatomy of childhood disruptive behaviordisorders. Aggress Behav. 2011; 37:1–12.

Whole Brain Conduct Disorder

PLOS ONE | DOI:10.1371/journal.pone.0155475 June 6, 2016 12 / 16

13. Herpertz SC, Huebner T, Marx I, Vloet TD, Fink GR, Stoecker T, et al. Emotional processing in maleadolescents with childhood-onset conduct disorder. J Child Psychol Psychiatry. 2008; 49(7):781–91.doi: 10.1111/j.1469-7610.2008.01905.x PMID: 18598245

14. Jones AP, Laurens KR, Herba CM, Barker GJ, Viding E. Amygdala hypoactivity to fearful faces in boyswith conduct problems and callous-unemotional traits. Am J Psychiatry. 2009; 166(1):95–102. doi: 10.1176/appi.ajp.2008.07071050 PMID: 18923070

15. Marsh AA, Finger EC, Mitchell DG, Reid ME, Sims C, Kosson DS, et al. Reduced amygdala responseto fearful expressions in children and adolescents with callous-unemotional traits and disruptive behav-ior disorders.[see comment][erratum appears in Am J Psychiatry. 2008 Jul;165(7):920]. Am J Psychia-try. 2008; 165(6):712–20. doi: 10.1176/appi.ajp.2007.07071145 PMID: 18281412

16. Passamonti L, Fairchild G, Goodyer I, Hurford G, Hagan CC, Rowe JB, et al. Neural abnormalities inearly-onset and adolescence-onset conduct disorder. Arch Gen Psychiatry. 2010; 67(7):729–38. doi:10.1001/archgenpsychiatry.2010.75 PMID: 20603454

17. Sterzer P, Stadler C, Krebs A, Kleinschmidt A, Poustka F. Abnormal neural responses to emotionalvisual stimuli in adolescents with conduct disorder. Biol Psychiatry. 2005; 57(1):7–15. PMID: 15607294

18. Rubia K, Smith AB, Halari R, Matsukura F, Mohammad M, Taylor E, et al. Disorder-specific dissociationof orbitofrontal dysfunction in boys with pure conduct disorder during reward and ventrolateral prefrontaldysfunction in boys with pure ADHD during sustained attention. Am J Psychiatry. 2009; 166(1):83–94.doi: 10.1176/appi.ajp.2008.08020212 PMID: 18829871

19. Decety J, Michalska KJ, Akitsuki Y, Lahey BB. Atypical empathic responses in adolescents withaggressive conduct disorder: a functional MRI investigation. Biol Psychol. 2009; 80(2):203–11. doi: 10.1016/j.biopsycho.2008.09.004 PMID: 18940230

20. Craig MC, Catani M, Deeley Q, Latham R, Daly E, Kanaan R, et al. Altered connections on the road topsychopathy. Mol Psychiatry. 2009; 14(10):946–53. doi: 10.1038/mp.2009.40 PMID: 19506560

21. Horsfield MA, Jones DK. Applications of diffusion-weighted and diffusion tensor MRI to white matter dis-eases—a review. NMR in biomedicine. 2002; 15(7–8):570–7. PMID: 12489103

22. Beaulieu C. The biological basis of diffusion anisotropy. In: Johansen-Berg HB, T. E. J., editor. DiffusionMRI: from quantitative measurement to in vivo neuroanatomy. London: Elsevier; 2009.

23. Paus T. Growth of white matter in the adolescent brain: myelin or axon? Brain Cogn. 2010; 72:26–35.doi: 10.1016/j.bandc.2009.06.002 PMID: 19595493

24. Sarkar S, Craig MC, Catani M, Dell'acqua F, Fahy T, Deeley Q, et al. Frontotemporal white-mattermicrostructural abnormalities in adolescents with conduct disorder: a diffusion tensor imaging study.Psychol Med. 2013; 43(2):401–11. doi: 10.1017/S003329171200116X PMID: 22617495

25. Smith SM, Jenkinson M, Johansen-Berg H, Rueckert D, Nichols TE, Mackay CE, et al. Tract-basedspatial statistics: Voxelwise analysis of multi-subject diffusion data. Neuroimage. 2006; 31:1487–505.PMID: 16624579

26. Sundram F, Deeley Q, Sarkar S, Daly E, Latham R, Craig M, et al. White matter microstructural abnor-malities in the frontal lobe of adults with antisocial personality disorder. Cortex. 2012; 48(2):216–29.doi: 10.1016/j.cortex.2011.06.005 PMID: 21777912

27. Hoppenbrouwer SS, Nazeri A, de Jesus DR, Stirpe T, Felsky D, Schutter DJLG, et al. White matter defi-cits in psychopathic offenders and correlation with factor structure. PLoS ONE. 2013; 8(8):8.

28. Zhang J, Zhu X, Wang X, Gao J, Shi H, Huang B, et al. Increased structural connectivity in corpus callo-sum in adolescent males with conduct disorder. J Am Acad Child Adolesc Psychiatry. 2014; 53(4):466–75. doi: 10.1016/j.jaac.2013.12.015 PMID: 24655656

29. Passamonti L, Fairchild G, Fornito A, Goodyer IM, Nimmo-Smith I, Hagan CC, et al. Abnormal Anatomi-cal Connectivity between the Amygdala and Orbitofrontal Cortex in Conduct Disorder. PLoS ONE.2012; 7(11):e48789. doi: 10.1371/journal.pone.0048789 PMID: 23144970

30. Caron-Haney E, Caprihan A, Stevens MC. DTI-measured white matter abnormalities in adolescentswith conduct disorder. J Psychiatr Res. 2014; 48(1):111–20. doi: 10.1016/j.jpsychires.2013.09.015PMID: 24139595

31. Finger EC, Marsh A, Blair KS, Majestic C, Evangelou I, Gupta K, et al. Impaired functional but pre-served structural connectivity in limbic white matter tracts in youth with conduct disorder or oppositionaldefiant disorder plus psychopathic traits. Psychiatry Res. 2012; 202(3):239–44. doi: 10.1016/j.pscychresns.2011.11.002 PMID: 22819939

32. Lenroot RK, Gogtay N, Greenstein DK, Wells EM, Wallace GL, Clasen LS, et al. Sexual dimorphism ofbrain developmental trajectories during childhood and adolescence. Neuroimage. 2007; 36(4):1065–73. PMID: 17513132

33. Zhang J, Gao J, Shi H, Huang B, Wang X, Situ W, et al. Sex differences of uncinate fasciculus structuralconnectivity in individuals with conduct disorder. BioMed research international. 2014; 2014.

Whole Brain Conduct Disorder

PLOS ONE | DOI:10.1371/journal.pone.0155475 June 6, 2016 13 / 16

34. Ingalhalikar M, Smith A, Parker D, Satterthwaite TD, Elliott MA, Ruparel K, et al. Sex differences in thestructural connectome of the human brain. Proceedings of the National Academy of Sciences. 2014;111(2):823–8.

35. Oldfield RC. The assessment and analysis of handedness: The Edinburgh inventory. Neuropsycholo-gia. 1971; 9(1):97–113. PMID: 5146491

36. Wechsler D. Wechsler Abbreviated Scale of Intelligence. Pearson Education, Inc; 1999.

37. Goodman R. The extended version of the strengths and difficulties questionnaire as a guide to childpsychiatric cases and consequent burden. Journal of child psychology and Psychiatry. 1999; 40:779–99.

38. Frick PJ, Hare RD. The Antisocial Process Screening Version (ASPD). Multi-Health Systems Inc.,Toronto, Canada. 2001.

39. Kaufman J, Birhamer B, Brent D, Rao U, Flynn C, Moreci P, et al. Schedule for Affective Disorders andSchizophrenia for school-age children—Present and Lifetime version (K-SADS-PL): initial reliabilityand validity data. J Am Acad Child Adolesc Psychiatry. 1997; 36:980–8. PMID: 9204677

40. Forth AE, Kosson D, Hare RD. Hare Psychopathy Checklist: Youth Version (PCL:YV). Multi-HealthSystems Inc., Toronto, Canada. 2003.

41. Finger EC, Marsh AA, Mitchell DG, Reid ME, Sims C, Budhani S, et al. Abnormal ventromedial prefron-tal cortex function in children with psychopathic traits during reversal learning. Arch Gen Psychiatry.2008; 65(5):586–94. doi: 10.1001/archpsyc.65.5.586 PMID: 18458210

42. Jones DK, Williams SCR, Gasston D, Horsfield MA, Simmons A, Howard R. Isotropic resolution diffu-sion tensor imaging with whole brain acquisition in a clinically acceptable time. Hum Brain Mapp. 2002;15:216–30. PMID: 11835610

43. Leemans A, Jeurissen B, Sijbers J, Jones DK. ExploreDTI: a graphical toolbox for processing analyz-ing, and visualizing diffusion MR data. Proceedings of the International Society for Magnetic Reso-nance in Medicine. 2009; 17(2):3537.

44. Chang LC, Jones DK, Pierpaoli C. RESTORE: robust estimation of tensors by outlier rejection. MagnReson Med. 2005; 53(5):1088–95. PMID: 15844157

45. Smith A, Jenkinson M, Woolrich MW, Beckmann CF, Behrens TEJ, Johansen-Berg H, et al. Advancesin functional and structural MR image analysis and implementation as FSL. Neuroimage. 2004; 23((S1)):208–19.

46. Catani M, Thiebaut de Schotten M. A diffusion tensor imaging tractography atlas for virtual in vivo dis-sections. Cortex. 2008; 44(8):1105–32. doi: 10.1016/j.cortex.2008.05.004 PMID: 18619589

47. Mori S, Wakana S, van Zijl PCM, Nagae-Poetscher LM. MRI atlas of human white matter. Amsterdam:Elsevier; 2005.

48. Craig MC, Catani M, Deeley Q, Latham R, Daly E, Kanaan R, et al. Altered connections on the road topsychopathy. Mol Psychiatry. 2009; 14(10):946–53, 07. doi: 10.1038/mp.2009.40 PMID: 19506560

49. Courchesne E, Karns CM, Davis HR, Ziccardi R, Carper RA, Tigue ZD, et al. Unusual brain growth pat-terns in early life in patients with autistic disorder. Neurology. 2001; 57:245–54. PMID: 11468308

50. Cheung C, Chua S, Cheung V, Khong P, Tai K, Wong T, et al. White matter fractional anisotrophy differ-ences and correlates of diagnostic symptoms in autism. Journal of Child and Adolescent Psychologyand Psychiatry. 2009; 50(9):1102–12.

51. Weinstein M, Ben-Sira L, Levy Y, Zachor DA, Itzhak EB, Artzi M, et al. Abnormal white matter integrityin young children with autism. Wiley Subscription Services, Inc., A Wiley Company; 2010.

52. Catani M, Thiebaut de Schotten M. Atlas of human brain connections. UK: Oxford University Press;2012. 544 p.

53. Berns GS, Moore S, Capra CM, Berns GS, Moore S, Capra CM. Adolescent engagement in dangerousbehaviors is associated with increased white matter maturity of frontal cortex. PLoS ONE. 2009; 4(8):e6773. doi: 10.1371/journal.pone.0006773 PMID: 19707512

54. Deoni SCL, Mercure E, Blasi A, Gasston D, Thomson A, Johnson M, et al. Mapping infant brain myeli-nation with magnetic resonance imaging. J Neurosci. 2011; 31(2):784–91. doi: 10.1523/JNEUROSCI.2106-10.2011 PMID: 21228187

55. Paus T, Zijdenbos A, Worsley K, Collins DL, Blumenthal J, Giedd JN, et al. Structural maturation of neu-ral pathways in children and adolescents: in vivo study. Science. 1999; 258:1908–11.

56. Berntson GG. Attack, grooming, and threat elicited by stimulation of the pontine tegmentum in cats.Physiol Behav. 1973; 11:91–7.

57. De Bellis MD, Narasimhan A, Thatcher DL, Keshavan MS, Soloff P, Clark DB. Prefrontal cortex, thala-mus, and cerebellar volumes in adolescents and young adults with adolescent-onset alcohol use disor-ders and comorbid mental disorders. Alcohol Clin Exp Res. 2005; 29(9):1590–600. PMID: 16205359

Whole Brain Conduct Disorder

PLOS ONE | DOI:10.1371/journal.pone.0155475 June 6, 2016 14 / 16

58. Nicolas JM, Fernandez-Sola J, Antunez RE, Cofan M, Cardenal C, Sacanella E, et al. High ethanolintake and malnutrition in alcoholic cerebellar shrinkage. QJM. 2000; 93(7):449–56. PMID: 10874054

59. Pfefferbaum A, Sullivan EV, Mathalon H, Lim KO. Frontal lobe volume loss observed with magnetic res-onance imaging in older chronic alcoholics. Alcohol Clin Exp Res. 1997; 21(3):521–9. PMID: 9161613

60. Sullivan EV. Compromised pontocerebellar and cerebellothalamocortical systems: speculations ontheir contributions to cognitive and motor impairment in nonamnesic alcoholism. Alcohol Clin Exp Res.2003; 27(9):1409–19. PMID: 14506401

61. Cadoret RJ, Yates WR, Troughton E, Woodworth g, Stewart MA. Adoption study demonstrating twogenetic pathways to drug abuse. Arch Gen Psychiatry. 1995; 52(1):42–52. PMID: 7811161

62. Lynskey MT, Fergusson DM. Childhood conduct problems, attention deficit behaviors, and adolescentalcohol, tobacco, and illicit drug use. J Abnorm Child Psychol. 1995; 23(3):281–302. PMID: 7642838

63. Ramnani N, Behrens TEJ, Johansen-Berg H, Richter MC, Pinsk MA, Andersson JLR, et al. The evolu-tion of prefrontal, inputs to the cortico-pontine system: diffusion imaging evidence frommacaquemon-keys and humans. Cereb Cortex. 2006; 16(6):811–8. PMID: 16120793

64. Tomasch J. The numerical capacity of the human cortico-pontocerebellar system. Brain Res. 1969; 13(3):476–84. PMID: 5305681

65. Schmahmann JD, Sherman JC. The cerebellar cognitive affective syndrome. Brain Cogn. 1998;121:561–79.

66. Stoodley CJ, Schmahmann JD. Functional topography in the human cerebellum: a meta-analysis ofneuroimaging studies. Neuroimage. 2009; 44(2):489–501. doi: 10.1016/j.neuroimage.2008.08.039PMID: 18835452

67. Turner BM, Paradiso S, Marvel CL, Pierson R, Boles Ponto LL, Hichwa RD, et al. The cerebellum andemotional experience. Neuropsychologia. 2007; 45:1331–41. PMID: 17123557

68. Sacchetti B, Scelfo B, Strata P. Cerebellum and emotional behavior. Neuroscience. 2009; 162:756–62.doi: 10.1016/j.neuroscience.2009.01.064 PMID: 19409218

69. Catani M, Jones DK, Daly E, Embiricos N, Deeley Q, Pugliese L, et al. Altered cerebellar feedback pro-jections in Asperger syndrome. Neuroimage. 2008; 41:1184–91. doi: 10.1016/j.neuroimage.2008.03.041 PMID: 18495494

70. Okugawa G, Nobuhara K, Minami T, Takase K, Sugimoto T, Saito Y, et al. Neural disorganization in thesuperior cerebellar peduncle and cognitive abnormality in patients with schizophrenia: a diffusion ten-sor imaging study. Prog Neuropsychopharmacol Biol Psychiatry. 2006; 30(8):1408–12. PMID:16844275

71. Klaas S, Friston KJ, Frith CD. Dysconnection in schizophrenia: from abnormal synaptic plasticity to fail-ures of self-monitoring. Schizophr Bull. 2009; 35(3):509–27. doi: 10.1093/schbul/sbn176 PMID:19155345

72. Schutter DJG, Hofman D, Van Honk J. Fearful faces selectively increase corticospinal motor tract excit-ability: a transcranial magnetic stimulation study. Psychophysiology. 2008; 45(3):345–8. doi: 10.1111/j.1469-8986.2007.00635.x PMID: 18221448

73. Birbaumer N, Veit R, Lotze M, Erb M, Hermann C, GroddW, et al. Deficient fear conditioning in psy-chopathy: a functional magnetic resonance imaging study. Arch Gen Psychiatry. 2005; 62(7):799–805.PMID: 15997022

74. Deeley Q, Daly E, Surguladze S, Tunstall N, Mezey G, Beer D, et al. Facial emotion processing in crimi-nal psychopathy. Preliminary functional magnetic resonance imaging study. Br J Psychiatry. 2006;189:533–9. PMID: 17139038

75. Kiehl KA, Smith AM, Hare RD, Mendrek A, Forster BB, Brink J, et al. Limbic abnormalities in affectiveprocessing by criminal psychopaths as revealed by functional magnetic resonance imaging. Biol Psy-chiatry. 2001; 50(9):677–84. PMID: 11704074

76. Makris N, Kennedy DN, McInerney S, Gregory Sorensen A, Wang R, Caviness VSJ, et al. Segmenta-tion of subcomponents within the superior longitudinal fascicle in humans: a quantitative, in vivo, DT-MRI study. Cereb Cortex. 2005; 15(6):854–69. PMID: 15590909

77. Jang SH, Hong JH. The anatomical characteristics of superior longitudinal fasciculus 1 in human brain:diffusion tensor tractography study. Neurosci Lett. 2011; 506(1):146–8. doi: 10.1016/j.neulet.2011.10.069 PMID: 22085696

78. Donnellan MB, Ge X, Wenk E. Cognitive abilities in adolescent-limited and life-course-persistent crimi-nal offenders. J Abnorm Psychol. 2000; 109(3):396–402. PMID: 11016109

79. Marcus DK, Lilienfeld SO, Edens JF, Poythress NG. Is antisocial personality disorder continuous or cat-egorical? A taxonomic analysis. Psychol Med. 2006; 36(11):1571–81. PMID: 16836795

Whole Brain Conduct Disorder

PLOS ONE | DOI:10.1371/journal.pone.0155475 June 6, 2016 15 / 16

80. Eluvathingal TJ, Chugani HT, Behen ME, Juhasz C, Muzik O, Maqbool M, et al. Abnormal brain con-nectivity in children after early severe socioemotional deprivation: a diffusion tensor imaging study.Pediatrics. 2006; 117:2093–100. PMID: 16740852

81. Choi J, Jeong BS, RohanML, Polcari AM, Teicher MH. Preliminary evidence for white matter tractabnormalities in young adults exposed to parental verbal abuse. Biol Psychiatry. 2009; 65:227–34. doi:10.1016/j.biopsych.2008.06.022 PMID: 18692174

82. Bengetsson SL, Nagy Z, Skare S, Forsman L, Forssberg H, Ullen F. Extensive piano practicing hasregionally specific effects on white matter development. Nat Neurosci. 2005; 8(9):1148–50. PMID:16116456

83. Tang Y-Y, Lu Q, Geng X, Stein EA, Yang Y, Posner MI. Short-term meditation induces white matterchanges in the anterior cingulate. Proceedings of the National Academy of Sciences. 2010; 107(35):15649–52.

84. Toritsuka M, Makinodan M, Kishimoto T. Social Experience-Dependent Myelination: An Implication forPsychiatric Disorders. Neural Plas. 2015; 2015:6.

85. Raine A, Moffitt TE, Caspi A, Loeber R, Stouthamer-Loeber M, Lynam D. Neurocognitive impairmentsin boys on the life-course persistent antisocial path. J Abnorm Psychol. 2005; 114(1):38–49. PMID:15709810

86. Goodman R, Simonoff E, Stevenson J. The impact of child IQ, parent IQ and sibling IQ on child beha-vioural deviance scores. Journal of Child Psychology and Psychiatry. 1995; 36(3):409–25. PMID:7782405

87. Wang Y, Adamson C, YuanW, Altaye M, Rajagopal A, Byars AW, et al. Sex differences in white matterdevelopment during adolescence: a DTI study. Brain Res. 2012; 1478:1–15. doi: 10.1016/j.brainres.2012.08.038 PMID: 22954903

88. Descoteaux M, Deriche M. Mapping neuronal fiber crossings in the human brain. Biomedical Opticsand Medical Imaging. 2008.

89. Jones DK. Studying connections in the living human brain with diffusion MRI. Cortex. 2008; 44:936–52.doi: 10.1016/j.cortex.2008.05.002 PMID: 18635164

90. Wolff JJ, Gu H, Gerig G, Elison JT, Styner M, Gouttard S, et al. Differences in white matter fiber tractdevelopment present from 6 to 24 months in infants with autism. Am J Psychiatry. 2012; 169(6):589–600. doi: 10.1176/appi.ajp.2011.11091447 PMID: 22362397

91. Maier-Hein KH, Brunner R, Lutz K, Henze R, Parzer P, Feigl N, et al. Disorder-Specific White MatterAlterations in Adolescent Borderline Personality Disorder. Biol Psychiatry. 2013.

92. Dell'Acqua F, Simmons A, Williams SC, Catani M. Can spherical deconvolution provide more informa-tion than fiber orientations? Hindrance modulated orientational anisotropy, a true‐tract specific index tocharacterize white matter diffusion. Hum Brain Mapp. 2013; 34(10):2464–83. doi: 10.1002/hbm.22080PMID: 22488973

93. Fortier CB, Leritz EC, Salat DH, Lindemer E, Maksimovskiy AL, Shepel J, et al. Widespread effects ofalcohol on white matter microstructure. Alcohol Clin Exp Res. 2014; 38(12):2925–33. doi: 10.1111/acer.12568 PMID: 25406797

Whole Brain Conduct Disorder

PLOS ONE | DOI:10.1371/journal.pone.0155475 June 6, 2016 16 / 16